.
_,--..
RECEI~EU
BEFORE THE
POSTAL RATE COMM[SSlON
WASHINGTON, D.C. 20268-ooo~~:~~~~~~E,,~~~~~~~
SPECIAL SERVICI:S REFORM, 1996
Docket No. MC96-3
IRESPONSE OF UNITED STATES POSTAL SERVICE
WITNESS ELLARD TO INTERROGATORIES
OF
THE OFFICE OF THE CONSUMER ADVOCATE
(OCAIUSPS-TG-7-12)
The United States Postal Service hereby provides responses of ,witness Ellard to
the following interrogatories
T6-7-12,
,,.-.
of the Office of the Consumer Advocate:
OCA/USPS-
filed on July 8, 1996. The original of these faxed respons’es together with
a signature page and USPS Library Reference SSR-128 will be tiled as soon ;as they
are received.
Tlhese items are apparently still en route from New Jersey.
Each interrogatory
is stated verbatim and is followed by the response.
Respectfully
submitted,
UNITED STATES POSTAL SERVICE
By its attorneys:
Daniel J. Foucheaux, Jr.
Chief Counsel, Ratemaking
&@..A9, &I&
Kenneth N. Hollies
475 CEnfant Plaza West, S.W.
Washington, D.C. 20260-1145
(202)268-3083;
Fax-5402
July 22, 1996
-_
OCAJUSPS-V-7.
Please refer to the file POBOX.DAT of SSR-I 11.
a.
Please confirm that either the weights supplied on file POBOX.DAT of SSR-111 are 106
times too Iergc or that the tabulations provided in Tables 2-7 are approximately
100
If
you
do
not
confirm,
please
explain
how
the
weights
provided
on
times too ?,mall.
POBOX.DAT should be used to tabulate survey data. If you do confirm. please explain
the reasons for the disparity.
b.
Please explain why the sum of weights for the first four observations of POBO:X.DAT is
about 99762, approximately three times as large as the total Group ‘I weighted base of
32988 as shown in Table 2 of your testimony.
RESPONSE to OGAIUSPS-TG-7.
a.
It is a common practice in the production of research tables to divide by 1100or 1,000 or
even by one million when representing weighted values.
tables more easily into available space.
This permits us to fit
The actual computations
the
use the ~full base.
Since all of the data reported in my testimony are in terms of percentage!;,
not absolute
numbers, this has no effect on that testimony
b.
Following my answer to Question 7a, divide 99,762 by 100 to see its true relationship to
the reported Group 1 weighted base of 32,988 as shown in Table 2 of my testimony.
2
Docket No. MC963
f-,
QCA’USPS-T69.
Please refer to Table 2 at page 53 of SSR-111
a.
Please confirm that the source of these figures was the PO Box Study described in
USPS-T-4.
If you do not confirm, please provide the source of this data. If you do
confirm, please provide a citation to the where these figures are present,& in witness
Lion’s testimony or library references.
b.
Please explain the large dismpancy baWeen the total weighted base of all respondents
(149,930’) in your testimony and the total numbar of post office boxes installed
(14,290.298) as reported by witness Lion in Table 3 of USPST4.
RESPONSE to OCAJUSPS-TG-6.
a.
This conilrms that the source of the post office box population figures cited on page 53
of SSR-11’1 was the P.O. Box Study described in USPS-T-4.
in USPS-T4
b.
r-”
The figures are presented
in Table 14 on page 37
As noted ih my response to Questions 7a and 7b.
the apparent discrepancy
cemes
form dividing the actual numbers by 100 for ease of presentation.
’ This is ,the sum of the weighted base for group 1 and the weighted base for group 2 as
shown in Table 12of USPS-T&.
Docket No. MC96-3
3
.r,--7
Please refer to page 17 of SSR-111. Question la asks “(Do you reside I
OCPJUSPS-TG-9.
Is your primary place of business) in the same ZIP Code as the towrl where you obtain
box service?”
Suppose a respondent’s residence ZIP Code is 20016 and they live on New Mexico Ave. NW,
Washington DC. This person does not rent a PO box at the closest loNation (Friendship,
20016). but instead rents a PO box at L’Enfant Plaza, SW, Washington DC: 20026. Suppose
that this box is used only for personal use.
a.
Please confirm that the residence ZIP Code is not the same as the ZIP Code where the
box was obtained. If you do not contirm, please explain.
b.
Please confirm that both the residence and the town where the box was: obtained are
Washingto;n, DC. If you do not confirm, please explain.
C.
Please explain what the correct response to question la would be for this respondent.
RESPONSE to OCANSPS-TG-9.
,,.--.
a.
Assuming that your data are correct, this is confirmed.
b.
Assuming that your data are correct, this is wnfinned.
C.
I’d say the correct answer is no. However, it is important to note that th’e respondents’
impression of the correct answer is far more important than mine. The purplose of the
question was to gain an understanding of the proportion of box hboldanj who believed
their boxes were in a ZIP Code other than that of their home address. To pin this
down specifically would require a more detailed question sequence.
4
Docket No. MC963
OCAIUSPS-TG-IO.
Please refer to page 18 of SSR-I 11. Question 3 asks, “If the fee: were to
be changed to B(MID-PRICE) for m,
would you accept it as something that cannot be
avoided or would you try to find an alternative?”
a.
Please confirm that the classification changes proposed for post office box. fees in
MlC96-3 are not restricted to a 6 month period.
b.
Were respondents informed that any box rent increases would not revert to current
prices after 6 months elapsed? Please explain,
C.
Would you anticipate a different reaction from respondents to this question if the
question did not restrict the increases to a 6 month period of time? For example,
suppose the question limited the increases to a 6 week period. Suppose the question
stated that rates would not be main
for at least 6 months. How could subtle
questionnaire wording changes such as these effect responses? Please explain.
d.
When you designed the questionnaire, did you understand that tested rates were to be
temporary (for 6 months) and then revert to current rates? Or was it your understanding
that box rates would be raised and not increased again for at least 6 months’? Please
explain.
e.
Did any respondents inquire about what would happen after the 6 month period?
what responses were they given.
f.
YOU
,/--‘.
If so,
state in your testimony that respondents tend to overreact to plice increases.
IJSPS-T-6, page 7, lines 8-9. Was this questionnaire designed to minimize or reduce
overreaction to price increases?
If so, how could this be accomplished?
Please
explain.
RESPONSE to CCAAJSPS-T6-IO.
10-a.
The fee quoted was that for a six month period, a normal billing period for post ofFice
boxes.
I did not believe, and I cannot believe, that the respondents believed that a fee
increase would be effective for only six months.
10.b.
The respondents
were not informed that any fee increase would not revert to current
prices after 6 months elapsed. No further explanation is necessary
10.~.
It goes without saying that a fee should be for a specific period of timte. Therefore.
period of time should be specified
a
in the question. Fees for post office boxes are
5
Docket No. MC9643
generally for six months or for a year. We decided to use six months in all the questions
bo,th because it was a familiar period for many of the box holdem and because it did not
after special computational
challenges for those who are billed annually.
If I were to use an odd period such as the specified six weeks, I’d b’e asking people to
use a fraction of 6126 weeks to compare to the normal six months billing.
If we try to make a questionnaire into a test rather than a reasonable conversation,
we
lalse the interest of the respondent and. probably, his or her cooperation.
Generally,
subtle differences
in question wordings lead to only subtle differences
in
results.
,y-,.
My intent was to make the questions as unambiguous as possible and to avoid even the
appearance of trying to affect responses.
10.d.
At no time did I ever think that the increase would be temporary
10-e.
I have ncl information on which to base an answer to this question.
instruct&
Interviewers
to repeat the question wording if it is unclear to the respondlent. We do not
Irecord un,solicited comments.
10.f. The questions were worded to present the price increase in a clear, factual manner,
,,--..
are
Docket No MC96-3
6
Please refer to the Statement of Work at pages 2-3 and the ‘ques‘&ionnaire
OCAIUSPS-TG-1 I,
at pages 24-30 of SSR-111. These pages indicate that the post office box late research
included a study of potential box renters currently on waiting lists for boxes.
a.
PIlease confirm that data were collected from respondents on waitins lists in this study.
If you do not confirm, please explain why there are figures in the column marked
“Waiting Completes” in the sample disposition printout of pages 42-4Sl of SSR-1 Il.
b.
Please confirm that data and tabulations from the waiting list respondents were not
induded in SSR-111 or in your testimony. If you do not confirm. please provide page
references to the tabulations.
C.
Are the respondents on the waiting lists more likely to seek alternatives to post office
boxes when faced with box rent increases? Please provide any tabulations used to
support your response.
RESPONSE to OCAIUSPS-TG-11.
,,.-.
11 .a.
This confirms that data were collected from respondents on waiting lists
11 .b.
This confirms that data tabulations for the waiting list customers were not included in
SSR-1 Il.
This also confirms that the result?. of the study were not discussed
in my
testimony.
11 .c.
The data from waiting list respondents were sparse (87 completed interviews:) and I do
not believe they are a suitable basis for any inferences.
subject to final processing.
Accordingly,
they were not
Moreover, since the Postal Service determined to include no
fee proposal for waiting list customers in its Request, there are no Iplans to process the
data further,
Docket No. MC963
7
OCAIUSPS-T&12.,
Please refer to page 17 of SSR-111. Questions la and 1b appear to
attempt to differentiate between resident and nonresident box holders under Ithe tested rates.
a.
Please provide tabulations and graphs for Tables 3-8 of your testimony
resident and nonresident box holders.
separately for
b.
Do your data show differences in price sensitivity between the resident and nonresident
box holders?
C.
Does your study indicate that the nonresident rate is the same for lbox hlolders renting
thle box for personal use compared to those ranting for business use? Please explain
and support with survey data results.
RESPONSE to OCAIUSPS-TG-12.
lZ.a-c. The requested tabulations are being filed as Library Reference SSR-128,
I do not wish to draw inferences or interpretations from these data since the study was
/“-‘-
not designed
to produce reliable results regarding diierences
non-resident
box
participants
,,.-.
holders.
However,
by presenting
may assert and defend any inference they wish.
the
between resident
PEqlJeSted
and
tabulations,
CERTIFICATE
OF SERVICE
I hereby certify that I have this day served the foregoing doculnent
participants
of record in this proceeding
in accordance with section 12 of ,the Rules of
Practice.
/Ld/
7, &l/A --
Kenneth N. Hollies
475 CEnfant Plaza West, SW.
Washington, D.C. 20260-1145
July 22, 1996
,_e-.
upon all